7 Variational Autoencoders
7.1 Training
Variational encoders are separately trained for each bird.
Figure 7.1: The Calinski-Harabasz Index (ratio of between-cluster to within-cluster variance) plateaus before the reconstruction loss for bird 7358.
Figure 7.2: Input (left) and decoded (right) syllables.
Figure 7.3: Traversing the embedding space from the centroid of syllable āiā to each other syllable centroid.
7.2 Syllable Clustering
Bird 7358 (66-68 DPH) has relatively stable syllables and song syntax, while bird 6951 (59-63 DPH) has more variable syllables and syntax 8.1.

Figure 7.4: Syllable clusters from embedded dimensions.
Figure 7.5: UMAP projection of song trajectory with neuron spikes shown as dots.